论文部分内容阅读
文中提出了一个新的、稳定的光谱导数Kalman滤波紫外分光光度法同时定量分析方法,并且成功地将其应用于极难分辨的苯酚-邻氯苯酚和2,4二氯苯酚三组分混合体系的同时测量。选择分析光谱导数的原因是其不仅包含着吸光度,还包括在指定波长位置变化趋向等更多的信息量,这样更利于在吸收峰重叠的位置捕捉有较大差异的信号。利用Kalman滤波可以解决由光谱导数而引起的噪声放大问题,也可以过滤源于实验的噪声以及来自传递模型误差。在260~290 nm区间测量了30组浓度各自在1~10 mg.L-1范围的苯酚-邻氯苯酚和2,4二氯苯酚标准混合溶液紫外吸收光谱图。利用分段延伸高次多项式回归模拟求导准确获得吸光度导数,并且利用偏最小二乘法将其制作成Kalman滤波标准工作系数矩阵。通过随机线性离散系统的Kalman滤波器最优平滑计算,对混合体系中的各组分进行定量分析。回收实验显示出,光谱导数Kalman滤波分析法应用于本实验的极难分辨的三组分体系,得到了非常高的回收率,并且在整个实验范围内光谱导数Kalman滤波分析法具有非常好的稳定性。
In this paper, a new and stable method for the simultaneous quantitative determination of UV-Vis spectrophotometry and Kalman filter is proposed, and it is successfully applied to the three-component mixture of phenol-o-chlorophenol and 2,4-dichlorophenol Simultaneous measurement. The reason for choosing to analyze the spectral derivative is that it not only contains the absorbance, but also includes more information such as the trend of the change in the position of the specified wavelength, which is more advantageous for capturing a signal with a large difference in positions where absorption peaks overlap. Using Kalman filtering solves the problem of noise amplification caused by spectral derivatives, as well as experimentally derived noise and from model transfer errors. The UV absorbance spectra of 30 standard solutions of phenol-orthochlorophenol and 2, 4-dichlorophenol were measured in the range of 260 ~ 290 nm in the concentration range of 1 ~ 10 mg.L-1. The piecewise extended high order polynomial regression simulation was used to derive the derivative of absorbance accurately, and then it was made into the Kalman filter standard work factor matrix by partial least square method. Through the optimal smoothing calculation of the Kalman filter of random linear discrete system, the components in the mixed system are quantitatively analyzed. The recovery experiment shows that the spectral derivative Kalman filter method is applied to the extremely difficult three-component system in this experiment, and a very high recovery rate is obtained. The spectral derivative Kalman filter analysis method has a very good stability throughout the experimental range Sex.